Media Processing Activity Group

1
Media Processing Activity Group

• Alok De, Dave Lindbergh, Imre Varga

2
Mission Statement

• The IMTC Media Processing Activity Group acts as
  a voice of the multimedia communication industry
  on media (incl. speech/audio, video) coding and
  processing. It identifies issues and priorities
  relevant to the IMTC members. It strives to
  influence standardization and policy in this
  regard by recommending cost-effective,
  easy-to-use, and low-IPR media (incl. voice)
  processing entities for multimedia communication.

3
Slight change...

• We are in the process of widening the scope, by
  not concentrating only to speech/audio, but also
  to other media codecs and processing, such as
  video, still picture.
• The current plans and slides, however, still do
  represent a major emphasis on speech / audio
  coding issues.

4
Understanding Requirements

• Identify and characterize multimedia
  communication applications in terms of technical
  and policy requirements.
• Understand requirements from delay,
  fixed/variable bit-rate, algorithmic complexity,
  scalability, layering, subjective media (e.g.
  speech) quality and IPR perspectives.

5
Developing Position Statements

• Explore technical/policy issues with IMTC
  members.
• Establish a set of agreed positions that require
  attention of standardization organizations.
• Develop IMTC position statements for dialoging
  with ITU, ETSI and other standard bodies and
  activity groups.

6
Building Knowledge Base on audio codecs

• Make open request to everyone with a speech codec
  (standardized/proprietary) for submission of
  pertinent information including technical
  description to IMTC.
• Gather information in a specific format (a
  template will be circulated).
• Collect audio output files (e.g., in .wav
  format) corresponding to IMTC-provided input
  speech.
• Share structured information among IMTC members.

7
Performing Gap Analysis

• Determine characteristics of available speech
  coders against technical/policy requirements
  projected.
• Compile factual statements based on information
  gathered, and codec output.
• Expose gaps to encourage further research in
  deficit and/or new areas relevant to our
  industry.
• Draw standard bodies attention so as to address
  these aspects while standardizing.

8
Community Involvement

• Voice Coder activity group leaders.
• IMTC member organizations with interest in speech
  coder design/analysis and are willing to share
  information for our InfoBase.
• Task Force members for addressing specific
  issues.
• Invited experts opinion in gap analysis phase.

9
Specific Plan

• Make communication to ITU related to speech coder
  standardization and our industry needs (approval
  of draft needed).
• Understand need from members related to G.722.1,
  AMR, and other codecs and make liaisons to ITU.
• Build audio database and share with others speech
  coding literature.
• Introduce speech coding-related tests in Superop.

10
Concluding Remarks

• Get cost-effective and low-IPR voice coders
  stan-dardized for multimedia communication. To
  meet this target, IMTC voice coder AG will heed
  to IMTC members requirements, assimilate
  information about various standard/proprietary
  speech coders, formulate report with experts
  opinions, influence standard bodies factually on
  technology/policy issues, and encourage
  researchers to attack deficit/new areas.